Vision

AI needs a captain.

One captain at the wheel — a crew of AI agents

AI will not remove humans from complex operations. It will change their position — from running the loop themselves to engineering the harness of agents around it, human in the loop at every consequential decision. I want to build the products and interfaces that make this transition practical, controllable and valuable inside real organizations.

The new skill isn't building with AI. It's knowing where to point it.

BusinessEvery agent has to trace to an outcome — leverage, revenue or retention, not a demo.

TechnologyMulti-agent systems are graph engineering, not one long autonomous loop — routing, retries and human-in-the-loop checkpoints built into the harness; autonomy is a dial, not a switch.

MarketingI write about growth marketing for SMEs — the best AI products don't sell "AI," they sell the outcome, and let the agent be the how.

How I work, from ambiguity to deployment: UnderstandUnderstand users, operations and constraints; FrameIdentify the problem that actually deserves solving; DesignDefine the experience, the flows, and the level of autonomy; ShipBuild and put a first version into production, fast; LearnWatch it run in the field and improve the product.